Three Things You Can Do with Database Time Travel Queries
Maintaining application state with the DBOS Transact open source durable execution librarybrary (Python and TypeScript) and DBOS Cloud not only makes applications resilient to failure, but also enables database time travel. DBOS automatically logs application state changes, including the history of interactions with your PostgreSQL database (or any PostgreSQL wire protocol-compatible DBMS), and you can query that information, as if you were querying your database at a specific point in time.
As a running example, we use the DBOS Widget Store demo app, simulating the backend of an online store. The demo app is written in TypeScript, with the open source DBOS Transact library (to enable automatic application durability and state change history), and deployed to DBOS Cloud, which we use to show the time travel features below.
The first use-case for time travel is time travel queries: querying your application database as of a past point in time. For example, you can query what products you had in inventory as of 12 PM yesterday. This is useful for analysis and reporting, allowing you to see how your application data evolves over time.
